Today Phil Mannion returns to the Musings with a guest post on his recent paper on dinosaur diversity patterns and their relationships to latitude. Take it away Phil:
The presence of a latitudinal biodiversity gradient (LBG), whereby species richness is highest in the tropics and declines polewards, is a pervasive pattern affecting the majority of life on Earth today, and was recognised by early naturalists such as Charles Darwin and Alexander von Humboldt (whose foundation coincidentally partly supported the research outlined below). Despite its near ubiquity (on both land and in the sea), the causes of the gradient are less well established, with numerous hypotheses proposed over the last fifty or so years. Most of these have been refuted, leaving climate and the distribution of area as the two most likely causes. Understanding the cause and evolution of the gradient is vital to predicting biodiversity loss driven by present-day climate change and explaining geographical variation in biodiversity; as such the fossil record offers a unique perspective on this issue.
Previous work investigating the deep time LBG focused on marine invertebrates – these studies tended to find support for a modern type pattern throughout the Phanerozoic (approximately the last 530 million years). Little prior work has been carried out on terrestrial species, but the few studies to look at the deep time LBG on land found no evidence for a modern pattern until after the Eocene (approximately 30 mya).
Along with colleagues from the UK (Roger Benson, Paul Upchurch, Richard Butler and Paul Barrett) and USA (Matt Carrano), we looked at the LBG in Mesozoic dinosaurs (including birds). Using a number of different methods to mediate for sampling biases in the fossil record, we found no evidence for a modern type pattern at any point in the 160 million year evolutionary history of Mesozoic dinosaurs; instead we found dinosaur diversity to peak at palaeotemperate latitudes (30-60° North and South). The consistency of this result across analyses for different time slices indicates that this pattern is not controlled by climatic fluctuations – evidence suggests that the climatic gradient was “flatter” in the Mesozoic than today (i.e. there was less of a difference in temperature between tropical and temperate regions) – but was instead driven by the amount of available land area in latitudinal belts.
Residual dinosaur diversity after controlling for sampling, plotted against non-marine area (NM area) and palaeogeographical reconstructions for the Late Triassic (bottom), Jurassic (middle) and Cretaceous (top). From Mannion et al., 2012
Given that living birds conform to the modern day pattern, a significant change must have occurred at some point in the last 65 million years. Evidence from molecular phylogeny (and work on fossil insects) suggests that this change occurred at the end of the Eocene (34 mya), following the strengthening of the climatic gradient and an increase in seasonality. As such, there is no evidence for a modern type LBG on land before the last 30 million years.
Mannion, P. D., Benson, R. J. B., Upchurch, P., Butler, R. J., Carrano, M. T. and Barrett, P. M. 2012 (Published online). A temperate palaeodiversity peak in Mesozoic dinosaurs and evidence for Late Cretaceous geographical partitioning. Global Ecology and Biogeography, doi: 10.1111/j.1466-8238.2011.00735.x
Dave Hone's Archosaur Musings 
What’s the current thinking on the likely cause(s) of this strengthening of the climatic gradient? Decrease in CO2 levels? Alterations to heat transport mechanisms from orogeny/tectonics?
I guess that the increase in seasonality would be a consequence of decreasing temperature converting more of the oceans to ice and exposing more land but were there other factors?
I may have read somewhere that the equatorial/low latitude environments of the mesozoic were overly hot or dessicated.
Is there a possibility of sampling bias? At least here in South America, there are far more dinosaur localities in the austral countries than in the tropical ones.
(I think I should have read the paper first, because the authors must have addressed this particular issue, but what the hell, I´ll ask anyway)